What a public knowledge ledger is and why we built one

Every public video gets transcribed thousands of times. Ten thousand people watch a conference talk, four hundred of them pull the transcript, and four hundred separate copies of the same work get made, paid for, and then lost inside four hundred private folders.

That's the waste the Knowledge Ledger is aimed at.

What it's

The Knowledge Ledger is a public feed. When somebody drops a public link, the summary of that link gets recorded: the title, the topic, the core ideas, the quotes and the tags. Anybody can read it without an account. Anybody can save it into their own library and work on it from there.

You can browse it now. It isn't a demo. It's what people actually dropped.

What's in an entry, and what isn't

An entry holds the summary and the extracted material, plus a link back to the original source so you can verify anything against it.

It doesn't hold who dropped it. There's no author name, no profile, no upvote count and no comment section on the entry itself. This is deliberate. The moment a public feed has attribution and scores attached, the incentive shifts from recording useful things to performing, and the feed fills with whatever performs.

Private material never appears. Uploaded files don't go there. Anything from a source that requires a login doesn't go there. Only public links, because a summary of something nobody else can open helps nobody.

The quality gate

A public feed of automated summaries fails in one specific way: it fills with entries where the extraction didn't work. A video with no speech, a page that returned an error, a link that resolved to a login wall. Each of those produces a technically valid entry containing nothing.

So there's a floor. If the transcript comes back under forty words, nothing is recorded. If the generated title or summary matches the shapes that failures produce, nothing is recorded. Skipped runs are logged so the failure patterns can be looked at, and skipped runs aren't charged.

This isn't a quality judgement on the content. Plenty of thin videos are worth watching. It's a check that the extraction produced something real.

Why it's worth doing

Work stops being repeated. If a talk is already in the ledger, the second person to want it reads the summary instead of paying to regenerate it. On popular material this is most of the value.

It shows what people actually pull. Not what a recommendation feed decides to show. What somebody thought was worth extracting, which is a much higher bar than a click.

It gives new users something real to look at. Any tool like this has a cold start problem. An empty product with a paste box tells you nothing about whether the output is any good. A feed of actual output tells you immediately.

It records the reasoning, not the link alone. A bookmark says somebody found this. A ledger entry says what was in it.

Saving an entry into your own library

Reading the ledger is free. Saving an entry pulls it into your library with the full context attached, including the transcript when the source was a public link, so every tool works on it straight away rather than needing the work redone.

That's the part that makes reuse real. A summary you can read but not build on is a nicer bookmark. A summary that arrives with the underlying text is the actual output.

What it isn't

It isn't a social network. There are no followers and no feed ranking. If you want to discuss something, Discussions exist as a separate thing, started deliberately from an entry.

It isn't a search engine. It contains what people dropped, which is a small and uneven slice of everything.

It isn't permanent by default. Entries can be hidden or removed. Sources go dead, material gets posted that shouldn't have been, mistakes happen.

The objection

The obvious objection is that this is republishing somebody else's work. Worth taking seriously.

An entry is a summary with attribution and a link back to the source. It isn't a copy of the video and it isn't a substitute for it. Quotes are kept short and marked as quotes. The source is always linked, so the entry sends people to the original rather than away from it. Where a source owner objects, the entry comes down.

The line we hold is the one that has always applied to writing about other people's work: summarise freely, quote briefly, attribute always, link back.

Where it fits

For most people the ledger is a side effect. You drop a link because you want the summary for yourself, and the summary happens to be public because the source was public. The private version of the same habit, which is where the compounding actually happens, is described in watch less, keep more.

The point is only this: if the work has already been done on a public thing, nobody should have to do it again.